Cheese Dossier · CF-1.3
MiMo-V2.5-Pro
xiaomi · Cheese #29· 5 signals · specialty instruction · Proprietary
CHEESE INDEX
61.5
DEPTH
64.8
BALANCE
66.1
CONSENSUS
66.1
STABILITY
63.7
COVERAGE
62%
VALUE
47.15
I / O
$0.435 / $0.87
SPECIALTY
instruction +15
BEST FIT
long_horizon 62
深度雷達
實線＝本模型 · 虛線＝群體中位
芝士軸
coding
agent
horizon
Cheese Index = 可用軸中位數；Balance = 三軸分歧小。
能力細節
swe ★
gpqa
hle
coding
agent ★
knowledge ★
instruction ★
reasoning ★
部署建議
中段指數：適合成本敏感路徑。優先看 Pareto／Value；用深度雷達找專長匹配，再與更高指數模型 A/B。
鄰近差分
vs GPT-5.6 Luna
coding-13.7
agent+3.5
horizon+4.7
指數鄰近
- GPT-5.6 Luna 61.5
- Claude Opus 4.7 (Adaptive) 61.6
- GPT-5.4 61.3
- Muse Spark 1.2 61.1
- Ornith-1.5-397B 61.0
同組織
- MiMo-V2.5 43.7